Search notes:

Data exchange formats

Some data exchange formats are

Challenges with CSV and JSON

CSV and JSON are noteably quite prone to data corruption, mainly because these format have almost no notion for data types or domains (such as date and/or time). So, it's easy to insert a text such as hello world into a field that is supposed to store a date.

Scalability, parallel processing

CSV and JSON cannot (easily) be split into partitions that can be processed in parallel.

See also

The Python library datasets

Index